To investigate the feasibility of deep learning-assisted rapid silent PET/MR in pediatric applications.

This study aims to address the challenges faced by PET/MR in pediatric applications by leveraging new technologies such as flexible coils, silent sequences, and deep learning-based image reconstruction. The objective is to provide a rapid, silent, comfortable, and half-dose PET/MR imaging solution for pediatric patients, and to demonstrate its feasibility in this patient population.

Inclusion Criteria:

1. Children who have clinical symptoms and are suspected of having malignant tumors after being evaluated by clinical pediatricians and require 18F-FDG PET/MR examination to further clarify their conditions.
2. Signed and dated informed consent form
3. Commit to complying with the research procedures and cooperate in the implementation of the entire research process.
4. No gender restriction, aged between 0 and 18 years old.

Exclusion Criteria:

1. Patients with serious illnesses that researchers consider unsuitable for participation in this clinical study. Such as severe heart and lung failure, severe bone marrow suppression, severe liver and kidney dysfunction, etc.
2. Patients with severe claustrophobia，epileptic seizures, respiratory distress who cannot lie down, and patients with mental anomalies who are unable to cooperate.
3. Patients with metal implants such as pacemakers, artificial metal heart valves, and neurostimulators in their bodies.
4. Non-compliant person.

Inpatients and outpatients at Ruijin Hospital who meet the inclusion criteria.